Let’s Explore

spec(incar)

cols(
  yfips = col_double(),
  year = col_double(),
  fips = col_double(),
  state = col_character(),
  county_name = col_character(),
  total_pop = col_double(),
  total_pop_15to64 = col_double(),
  female_pop_15to64 = col_double(),
  male_pop_15to64 = col_double(),
  asian_pop_15to64 = col_double(),
  black_pop_15to64 = col_double(),
  latino_pop_15to64 = col_double(),
  native_pop_15to64 = col_double(),
  other_pop_15to64 = col_double(),
  white_pop_15to64 = col_double(),
  urbanicity = col_character(),
  region = col_character(),
  division = col_character(),
  commuting_zone = col_double(),
  metro_area = col_double(),
  land_area = col_double(),
  total_jail_adm = col_double(),
  total_jail_adm_dcrp = col_double(),
  female_jail_adm_dcrp = col_double(),
  male_jail_adm_dcrp = col_double(),
  total_jail_pop = col_double(),
  female_jail_pop = col_double(),
  male_jail_pop = col_double(),
  asian_jail_pop = col_double(),
  black_jail_pop = col_double(),
  latino_jail_pop = col_double(),
  native_jail_pop = col_double(),
  white_jail_pop = col_double(),
  total_jail_pretrial = col_double(),
  female_jail_pretrial = col_double(),
  male_jail_pretrial = col_double(),
  jail_from_state_prison = col_double(),
  jail_from_other_state_prison = col_double(),
  jail_from_state_jail = col_double(),
  jail_from_other_state_jail = col_double(),
  jail_from_fed = col_double(),
  jail_from_ice = col_double(),
  total_jail_pop_dcrp = col_double(),
  female_jail_pop_dcrp = col_double(),
  male_jail_pop_dcrp = col_double(),
  total_prison_pop = col_double(),
  female_prison_pop = col_double(),
  male_prison_pop = col_double(),
  asian_prison_pop = col_double(),
  black_prison_pop = col_double(),
  latino_prison_pop = col_double(),
  native_prison_pop = col_double(),
  other_prison_pop = col_double(),
  white_prison_pop = col_double(),
  total_prison_adm = col_double(),
  female_prison_adm = col_double(),
  male_prison_adm = col_double(),
  asian_prison_adm = col_double(),
  black_prison_adm = col_double(),
  latino_prison_adm = col_double(),
  native_prison_adm = col_double(),
  other_prison_adm = col_double(),
  white_prison_adm = col_double(),
  num_facilites = col_double(),
  num_employees = col_double(),
  confined_pop = col_double(),
  capacity = col_double(),
  ucr_population = col_double(),
  index_crime = col_double(),
  violent_crime = col_double(),
  property_crime = col_double(),
  murder_crime = col_double(),
  rape_crime = col_double(),
  robbery_crime = col_double(),
  agr_assault_crime = col_double(),
  burglary_crime = col_double(),
  larceny_crime = col_double(),
  mv_theft_crime = col_double(),
  arson_crime = col_double()
)

dim(incar)

[1] 147533     79

What missing data do I have?

Is missingness related to year?

incar %>% 
  group_by(urbanicity) %>% 
  summarise(
    min = min(total_prison_pop, na.rm = TRUE),
    median = median(total_prison_pop, na.rm = TRUE),
    max = max(total_prison_pop, na.rm = TRUE)
  )

Which counties have the highest prison population per capita

incar %>% 
  filter(year >= 2000, state == "TX") %>%
  inner_join(top_texas_counties) %>% 
  select(year, state, county_name, total_prison_pop, total_pop) %>% 
  mutate(prison_pop_per_capita = total_prison_pop / total_pop) %>% 
  ggplot(aes(x = year, y = prison_pop_per_capita, col = county_name)) +
  geom_line(show.legend = FALSE) +
  scale_y_continuous(labels = scales::percent) +
  facet_wrap(~county_name, scales = "free") +
  theme_bw()

Joining, by = "county_name"

Which county’s have the biggest change?

incar %>% 
  filter(year >= 2000) %>% 
  mutate(prison_pop_per_capita = total_prison_pop / total_pop) %>% 
  group_by(county_name, state) %>% 
  summarise(variance = var(prison_pop_per_capita, na.rm = TRUE),
            spread = max(prison_pop_per_capita, na.rm = TRUE) - min(prison_pop_per_capita, na.rm = TRUE))